On November 7, 2010, 3.3 million Brazilian secondary school graduates, hoping for a place in a university, took the National Assessment of Secondary Education (ENEM), a two-day examination marathon covering the humanities, natural sciences, language and mathematics. The next day, it appeared that in some places the answer sheets had not been printed correctly, leading to errors in test correction. A few days later, a federal judge suspended the exam, and ordered the Ministry of Education to do it over. The government appealed, and only those individuals directly affected will get the chance to take the test again in December. In 2009, the whole exam had to be annulled and done again after it was discovered that there were people selling the correct answers.
The Ministry of Education and its statistical and assessment office, the National Institute of Education Research (INEP), were strongly criticized for their mismanagement of the test, which had already led to the dismissal of INEP’s president after the 2009 crisis. More important than the management issue, however, is whether this assessment is useful in the way it is currently conceived.
ENEM was introduced in 1998 as a voluntary test for secondary school leavers, as an attempt to establish a standard of reference for the quality of the education they received: it was conceived as a “single, multidisciplinary test consisting of an essay and 63 objective questions, based on a matrix of five competencies and 21 abilities.” These competencies included fluency in Portuguese and the mathematical, artistic and scientific languages; the use of concepts and data in all fields of knowledge; the construction of consistent arguments; and the capacity to elaborate proposals to intervene in current social reality while respecting human values and taking into account the socio-cultural diversity of the country. 
The hope was that universities would use the test results in their admission process. By 2002, 338 institutions (mostly private) adhered to it, and 1.3 million students took the test. Public universities, however, that are more selective and more demanding, did not. In 2009, an agreement seemed to have been reached between the Ministry of Education and the federal universities— the test would become more demanding and content-based and, subsequently, the universities would give more weight to it in their admission procedures, either reserving a percentage of their places for admissions through ENEM or giving some weight to it along with other tests and admission requirements.
Besides its use in university admissions, the government published the mean scores of ENEM applicants from each secondary school in the country as a way of ranking them, and used it as one criterion to select students for a program of free tuition in private institutions – the PROUNI program.
ENEM has been criticized on many grounds. There is no evidence that it actually measures the competencies and abilities it pretends to. It probably measures only the student’s ability to read and understand the questions which is strongly correlated with the student’s family background and income levels. The current version separates items into four different broad subject areas and the language and humanities tests have been strongly criticized for their arbitrary and ideological biases. Since all students have to take the whole test, it forces all secondary schools to prepare the students the same way, not allowing for choice and differentiation, so much in need at this level. Finally, the most selective universities are not likely to replace their main admission procedures for such a test, even if they consider it to some extent only to please the authorities at the Ministry of Education.
The original idea, that Brazil’s secondary education needs external standards, and that the student’s competencies, as measured by these standards, could be used by the universities in their admission procedures, is still valid. But there should be multiple standards (in language, humanities, natural sciences, biological sciences, social sciences, arts) instead of one; these standards should be clear enough to allow the schools to understand their meaning and use them to improve their teaching; and higher education institutions should be able to chose which standards they think make more sense for the students they are looking for in different careers and levels. Finally, in the age of information technology and item response theory, there is no reason to cram 3 million students in a single two-days marathon test in thousands of locations, which can only lead to trouble.
Castro, Maria Helena Guimarães, and Sertio Tiezzi. 2004. The reform of secondary education and the implementation of ENEM in Brazil. In The Challenges of Education in Brazil, edited by C. Brock and S. Schwartzman. Oxford, UK: Triangle Journals, Ltd.